009c i - nasa€¦ · a diode element occurs in series with the edge channel. in the case of the...
TRANSCRIPT
NASA CR 114429da/ 'h Available to the Public
Analysis Report Volume II
SILICON SOLAR CELL DEVELOPMENT ANDRADIATION EFFECTS STUDY FOR LOWTEMPERATURE AND LOW ILLUMINATIONINTENSITY OPERATION
Contract NAS2-5516
(NASA-CR-114429) SILICON SOLAR: CELL N72-26033DEVELOPMENT AND RADIATION EFFECTS STUDY FORLOW TEMPERATURE AND LOW ILLUMINATIONINTENSITY A.-B. Kirkpatrick .(Ion Physics UnclasCorp.), Jan. 1972 49 ip, CSCL 10A G3/03 30812
Submitted by:
Allen R. Kirkpatrick
Submitted to:
National Aeronautics and SpaceAmes Research CenterMoffett Field, California
Technical Monitor: R. J. Debs
.TJanu1arvr 1 Q'72
Reproduced by
NATIONAL TECHNICALINFORMATION SERVICE
Li c, t....3 - -or ......o.rcecl .aLss4,a ,Y v, v 'r. u a DepartmenT ot Commerce
Springfield VA 22151
ION L ]DYz7 009c IA Subsidiary of High Voltage Engineering Corporation
BURLINGTON, MASSACHUSETTS
https://ntrs.nasa.gov/search.jsp?R=19720018383 2020-06-20T05:27:42+00:00Z
Analysis Report Volume II
SILICON SOLAR CELL DEVELOPMENT ANDRADIATION EFFECTS STUDY FOR LOW
TEMPERATURE AND LOW ILLUMINATIONINTENSITY OPERATION
Allen R. Kirkpatrick
National Aeronautics and Space AdministrationAmes Research Center
Moffett Field, California
Technical Monitor: R. J. Debs
Contract NAS2-5516
ION PHYSICSBURLINGTON,
CORPORATIONMASSACHUSETTS
PRECIDING PAGE BLANK NOT PT' V iiM
TABLE OF CONTENTS
Section Page
1 INTRODUCTION AND SUMMARY .................. 1
2 SUMMARY OF LOW TEMPERATURE/LOWINTENSITY EFFECTS ......................... 3
2.1 Performance Problems .................... 3
3 IRRADIATION STUDIES ........................ 8
3.1 Introduction ............................ 83.2 Experimental Study ....................... 83.3 Irradiation Test Data ...................... 11
3.3.1 In-Situ Irradiation and Measurement ....... 113.3.2 Performance at Temperatures Other
than that of Irradiation ................ 153.3.3 Annealing Effects ................... 28
3.4 Effect of Irradiation on the EquivalentCircuit Model ........................... 30
4 CELL SPECIFICATION ............ ......... 37
4.1 Cell Specification ........................ 38
5 TEST PLAN ................................ 42
6 REFERENC ES ............................... 45
ii
LIST OF ILLUSTRATIONS
Figure Page
1. Typical Low Temperature, Low Intensity Solar CellOutput Characteristics ........................ 4
2. Equivalent Circuit of the N/P Solar Cell ............. 5
3a. Illuminated I-V Characteristics at -160 °C asFunction of Irradiation Fluence ................... 12
3b. Illuminated I-V Characteristics at 28 °C asFunction of Irradiation Fluence ................... 13
4. The Effect of Irradiation on a Cell withExcessive Edge Conduction ...................... 14
5. The Effect of Temperature on Degradation of Pmax ..... 15
6. Isc Degradation at 28 °C ......................... 17
7. Isc Degradation at -70 °C ............. ........... 18
8. Isc Degradation at -160 °C ...................... 19
9. Isc Degradation at -180 °C ...................... 20
10. Pmax Degradation at 28 °C ...................... 21
11. Pmax Degradation at -70 °C ..................... . 22
12. Pmax Degradation at -160 °C ................... .. 23
13. Pmax Degradation at -180 °C ..................... 24
14. Variation of Output I-V Characteristic DuringIn-Situ Storage ............................... 29
15. The Effect of Irradiation on the DarkI-V Characteristic ........................... 33
16. Dark I-V Characteristic of Excessive EdgeConduction Cell ............................. 35
17. The Influence of Irradiation on the Solar CellCircuit Model ............................... 36
18. Representative Dark Linear I-VCharacteristics at 77 °K ........................ 43
iii
SECTION 1
INTRODUCTION AND SUMMARY
(1)Volume I( 1 ) of this Analysis Report described investigations performed
to identify and analyze the mechanisms responsible for anomalous performance
effects observed in silicon solar cells at low temperatures under low illumination
intensities. Performance deficiencies were categorized, mechanisms involved
were explained and necessary corrective steps in fabrication were outlined.
This Volume II of the Analysis Report describes completion of the
examination of performance problems and presents the results of a study to
determine the effect of in-situ proton irradiation upon low temperature, low
intensity performance of several cell types. Included in this report are recom-
mendations for an optimized cell for Jupiter probe use and definition of the
testing required on these cells to insure good performance characteristics.
Five cell types were included in the in-situ 1 MeV proton irradiation
study. These types were selected in an attempt to distinguish variations in
temperature-dependent radiation resistance which could be attributed to the
N/P or P/N structure, diffused or implanted junctions, crucible grown or float-
zone type base material and high or low base resistivity. While expected varia-
tions in performance were observed at room temperature, all cell types degraded
more or less similarly at lower temperatures with normalized degradation be-
coming increasingly rapid as temperature was reduced. A number of annealing
effects were observed but it is evident that any solar cell operating at low
temperature will be extremely vulnerable to radiation-induced degradation.
The rapid degradation of cell output at low temperatures due to proton
damage demands that cells be particularly well shielded against any proton en-
vironment associated with a low temperature mission. In the case of a Jupiter
probe there exists a strong possibility that high fluences of trapped protons with
energies in the range 5 to 50 MeV will be encountered. Solar cells to be used
on such a mission should be shielded against as much of this environment as
1
possible. A small area panel with exceptionally thick coverglass protection
used in conjunction with solar concentrators could be advantageous.
2
SECTION 2
SUMMARY OF LOW TEMPERATURE/LOWINTENSITY EFFECTS
2.1 Performance Problems
Classification and explanation of problems associated with solar cell
low temperature/low intensity operation have been discussed at length in
Volume I ( 1 ) of this Analysis Report. In brief, it is possible to identify cell
operation at low temperature and intensity as falling into one of four categories
illustrated in Figure 1. Another operational group has been reported elsewhere(2 )
as including cells which experience excessive Isc decrease as temperature is
reduced. The performance categories of Figure 1 include:
(1) 'ideal' high performance
(2) rectifying contact problem
(3) low shunt resistance problem
(4) 'double slope' edge conduction problem
Incidence of the operational problems is strongly dependent upon cell
fabrication procedures. Consequently the distribution of cells among the good
and problem categories has varied widely among conventional cells from different
manufacturers. Observed behavior can be explained in terms of an equivalent
circuit derived from a simple model of the solar cell structure.
Equivalent circuit elements which must be considered in association
with the cell junction are shown in Figure 2. The cell is separated into body
and edge regions. The body consists of the p-n junction, a junction defect shunt
resistance in parallel with the junction and possibly a Schottky barrier in series
with the junction at the back contact. The back contact Schottky barrier will not
be present if precautions are taken during fabrication to assure its absence.
The cell edge consists of a resistive element in series with a rectifying element.
More than one explanation for this edge model is possible. IPC believes
that the resistive element is formed by a thin inversion or accumulation
3
Ut
(VUI) I
1-6359
4
O
.~
co
C=
r-
CD
L) 0
a)
cq
O
.el
0-a)
--o'e
l v
'e
C
C;
0q
C;a)
r-4'-
C0
o
o
Channel].RCH IResistance I
N- Si/Metal IN- Si/Metal iSchottkyBarrieror NP-
'
Dio de |
EDGE I
Figure 2.
RS Series Resistance
'"'t I Shunt 2 PN Junctionenerated L Rurrent RJD
P-Si/MetalSchottky Barrier(not present ifP' layer is usedbelow metalli-zation)
BODY
Equivalent Circuit of the N/P Solar Cell.
1-5092
5
layer channel on the silicon surface, and the junction or Schottky barrier diode
resides at the interface between this channel and opposite type silicon or contact
metallization. The mechanisms involved have been described in detail in
Volume I of this report and in reference 3.
A cell exhibiting the characteristic rectifying contact problem be-
havior at low temperature is one in which a Schottky barrier is present in the
body branch of the cell equivalent circuit at the back contact. The Schottky
barriers which is produced at the interface between lightly doped base region
silicon and the contact metals is forward biased when the cell is illuminated.
Cell output voltage is lost in order to forward bias the barrier to conduct cell
current. The barrier can be eliminated by introducing a heavily-doped layer
into the silicon below the back contact by alloying or diffusion.
Directly across the cell junction is a shunt path with resistance which
is determined by the presence of junction defects. In a typical good cell this
resistance is very high and is of no consequence to cell performance even under
illumination intensities of only a few milliwatts per cm2. However, damaged
junction structures can be fabricated in which the junction shunt resistance is
sufficiently low to cause poor fill factors under low intensity operation. Gen-
erally this will occur because of penetration of the front contact metallization
through the junction, probably in the form of localized filamentary paths. The
problem can best be avoided by producing uniformly deep junctions not locally
shallower than 0 .2 5 g over the cell surface.
The presence of a conducting channel down the edge of the cell is of
considerable importance to the low temperature, low intensity cell operation.
A diode element occurs in series with the edge channel. In the case of the N/P
cell with N-type surface inversion layer, the rectifying element is formed at
the interface of the channel layer and the back contact metallization (or the
heavily doped P+ layer sometimes introduced under the contact). The edge
channel diode is always biased in the same direction as the cell junction. At
room temperature negligible voltage is required to pass forward current across
this diode and the edge component of the structure appears simply as a contribu-
tion to shunt resistance. At lower temperatures up to several hundred millivolts
6
may be required to turn-on the edge diode. At output voltages below that pro-
ducing diode turn-on, the diode blocks current through the edge channel. When
turn-on occurs, the edge channel appears as a conductive shunt between the cell
front and back contacts. If resistance of the edge channel is sufficiently low,
the 'double slope' problem output characteristic results.
Assured elimination of the excessive edge conduction problem would
require major modification to existing conventional cell structure. Such modifi-
cation could, for example, involve use of a planar geometry without exposed
junction at the cell edge. However IPC has observed only occasional incidence
of the double slope problem in cells sized by scribing after contacts and anti-
reflective coatings have been applied. Also it has been found that an etched
edge cell exhibiting the double slope problem can generally be improved by
re-etching its edges in a 6-1-1 (parts HF, HNO 3 , HAc) solution. Use of a
scribed edge cell and simplified dark liquid nitrogen submersion testing described
in Section 5 should adequately avoid the double slope performance deficiency.
7
SECTION 3
IRRADIATION STUDIES
3.1 Introduction
In order to define a cell design optimized for operation at low tem-
perature and intensity on a mission which could also include simultaneous
encounter with a significant radiation environment, some knowledge of degrada-
tion effects to be involved is required. At present in the absence of flight data,
it is not possible to anticipate with confidence the radiation threat which will be
experienced by a probe to Jupiter. Predictions which have been made(4, 5) of
the Jovian environment suggest that the electron component will probably be of
minor consequence compared to potential proton radiation exposure. The basis
for this conclusion is that protons in any Jovian trapped radiation belts are pre-
dicted to be of sufficiently high energy to penetrate coverglass protection of any
realistic thickness.
Existing experimental data related to solar cell irradiations at low
temperatures indicate electron exposure effects(6) to be more, but not exces-
sively more, serious at low temperature than at 28°C. On the other hand, proton
irradiations at low temperatures(7 ) have resulted in much greater relative
degradation than is observed at room temperature. Because of the importance
of the proton environment and the apparent strong temperature dependence of
proton degradation effects, it was decided to investigate proton degradation as
a function of irradiation temperature for a number of silicon cell types. Ex-
ceptional resistance to proton degradation at low temperatures by specific cell
types could be of considerable importance in the selection of an optimized solar
cell for a Jupiter mission.
3.2 Experimental Study
A proton irradiation program was carried out on cells of five types
described in Table I. Centralab cells were supplied by Dr. R. J. Debs of NASA
8
Ames Research Center. Cell leads, except those to the backs of the Centralab
P/N and N/P cells, were 0.070" x 0.003" aluminum strip attached by ultra-
sonic welding. Unreliable welds were obtained on some of the Centralab cell
backs and soldered silver strip was used instead. Redundant leads were used
to allow separate monitoring of cell current and voltage. A common back lead
was used for strings of three cells and the two ends of the lead served as the
redundant connections.
Cells to be irradiated were mounted in groups of six on 0. 250" thick
copper plates which were fastened to a coolant reservoir. Each group included
the five cell types of Table I. Slots in the plates accommodated the cell common
back leads which were clamped to the plates at the external lead connection posts.
A heated copper block used between the cell plate and reservoir allowed con-
tinuous variation over a wide temperature range but generally measurements
were made only at 28 °C, -70 °C, -160 °C and -180 °C. Temperature was moni-
tored using thermocouples attached to the cell plate. Irradiation series were
performed at specific constant temperatures which were maintained without
interruption throughout the irradiation and data accumulation sequences. Prior
to irradiation and after the final irradiation step of a series, data was taken at
other selected temperatures.
Irradiations were done using an IPC 2 MeV Van de Graaff accelera-
tor. A beam of protons of 1 MeV energy was electrostatically scanned to uni-
formly irradiate a 5 x 8 cm area of the sample holder. The fluence was deter-
mined from the total charge carried by the beam to the sample holder. Since
the mean range of 1 MeV protons in silicon is 16 micrometers, radiation damage
was produced to a depth corresponding to about 50 times the junction depth.
After each irradiation step the cell holder and reservoir could be rotated 180 °
away from the proton beam port to face illumination through a 6" diameter window.
Illumination was provided by an ultrastable 3200 °K color temperature unfiltered
tungsten-iodine source. Intensity over the cell holder was measured using an
Eppley thermopile and was monitored during each data recording. Intensity
varied by +2% over the cell holder and remained stable to within ± 1% during
irradiation runs.
9
a)
UO
.; Cl
U) -4
',.44 U
U U O U) u
Q * * **S
bfl bDt
Cd Cd 1~
o a E E
Q E H H _
0 In C) C> Ln L
Z_
o ; ; C;
oE E E" aa.4- +;+ +
_4 U) ) U)
C) C) U) U)S - *'- *
- U)+
Pi P4 -4 A4 .z
4j pi .pq p -4 -410 0- 0- 110 +
m ~ ~ ~ d
re 0 0 0 n
- - 1-4 -IH 1-4
Pi4 P.4 Al4
m n
0 --. -. ,- ka ~ + z P
c=~~~~~c ccc
c ~P. P.I P.c -P
'4-4 - a - a -
k ~ ~~ UP
'-. C~ J L) U U U ed ed~ t
HPP~lF
10
U)
U)U)
E-4
o
.4'-4
'-4-
Cl
e8
$-4
H
1-4
z
-4
En
U)PF-- 4
Ev-4
a)
UQ1-4a,CT
H
a,03
-4
CdE
av
;.4U)
SI
(3)
C)
'.I
-
'.4
a)
Q~ Sp4-4
. 4c
M
- o
* **
3.3 Irradiation Test Data
3.3.1 In-Situ Irradiation and Measurement
The effect of a series of low temperature irradiations on the in-situ
low temperature, low illumination intensity output I-V characteristics of a typical
cell is illustrated in Figure 3a. The results of equivalent irradiations and mea-
surements at 28 °C on a similar cell are shown for comparison in Figure 3b.
As cell operating temperature is decreased, irradiation causes more
severe degradation of output current and with one exception has only moderate
effect on Voc. The exception can occur in the case of a cell with the double slope
problem at low temperature and intensity. A rather extreme example is shown
by cell 10 CG84-9 of Figure 4. In the case of a cell with less serious double
slope deficiency, Voc would degrade slowly with increasing irradiation as long
as Isc exceeded edge current at Voc. When maximum edge current became
essentially equal to Isc, Voc would become limited by the double slope portion of
the I-V characteristic and would decrease rapidly with decreasing current.
Degradation of cell maximum power is primarily dependent upon
loss of output current and consequently also becomes increasingly rapid as
operating temperature is reduced. In Figure 5 normalized maximum power
output for a particular cell type is given as a function of proton fluence for
irradiation and measurement at 28 °C, -70 °C, -160 °C and -180 °C. The same
trend is observed for all cell types investigated.
A series of irradiations and measurements in-situ at 28 °C, -70 °C,-160 °C and -180 °C were made in which cells utilized were screened using the
procedures of Section 5 to eliminate those with bad initial performance at low
temperature and intensity. Because limited numbers of diffused CG 10 ohm-cm
N/P, implanted Lopex 1 ohm-cm N/P and diffused CG 10 ohm-cm P/N cells were
available, it was in some cases necessary to utilize cells with low shunt resis-
tance or minor double slope characteristics and eliminate only those with serious
deficiencies. However, all cells for these runs exhibited at least moderately
good output characteristics. Normalized low intensity Isc versus
11
0
o, ~o
cl- a)~~~4-
0 F:0.0o
0~~~~~~~~~~~~~~
Cd
CZ~~~~~~~~~~~~~~~~~~~~~~~c
O~ 0OJ I Ic
ouE~~~~ cc~ o
c c 0 1 lio
>
WI~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
cq~~~~~~~~~~~~~~~~c
CD CD~~~~~~~~~~
W~~~~~ 0 Q
0 CD 0
·· I I I ·Il
zI I I I)i
E E: I W .
0 o~~~~~0,
1-6318
12~~~~~~~~~~~e
O E .0
- a U)
C\1~~~~~~Y ~~0 0)coCC~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.
a;;;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ac'~~~~~~~i 0 ~ ~ ~ ~ c
0 0 0~~
xm *
a)
NE~~~~ 0
(v.u IIlr
1-6318~~~~~~~~~Q012 0 10F
6
5
4
3
2
1
0
0.1 0.2 0.3 0.4 0.5 0.6
V (volt)
Figure 3b. Illuminated I-V Characteristics at 28 °C as Functionof Irradiation Fluence.
1-6319
13
0.7
Initial
Cell 10 CG 84-9Diffused N/P 10Q-cm CGIrradiation and Measurement -
After 1.7 x 1011 p/cm2
0.1 0.2 0.3
185 C5.0 mW/cm
I0.4
V (volt)
Figure 4. The Effect of Irradiation on a Cell withExcessive Edge Conduction.
1-6320
14
5
4
3 -
2
1
00.5 0.6 0.7 0.8
C0I I ~ ~ ~~~I I o -I~~
iII\ U0 UC- D
I-;
//
rL4 E c
z
0 IIP40
_) -
0
-40 -4
cu 'd~~~~~~~~~~~o
o~ F:bil
~ 0
C~~~~~~~~~~~~~~~~~~~~C
1o~~~~~ 0 cIc
.E-q
I I
o
$ o Q,~~~~~~~~~~
0 4- c~ ~ ~ -
0~~~~~~~~~~~~~~7 X -4~~~~~~~~~~~~~~~~~~~~~~~~~~~-
0 ~~ (0(0C'
-4 E
I~u!*(x tud)/ x u d
1- 6321
15
proton fluence curves for individual cells of each type are given in Figures 6, 7,
8 and 9 for irradiation and measurement temperatures of 28 °C, -70 °C, -160 °C
and -180 °C respectively. Associated normalized Pmax curves for the same
cells are shown in Figures 10 through 13. It is apparent from inspection of
these curves that while there exists a wide variation in sensitivity of the five
cell types to proton irradiation at room temperature, at -70 °C and below, all
types exhibit similar, perhaps almost identical, degradation rates.
Fluences of 1 MeV protons to degrade cell Isc and Pmax by 25 and
50% at each irradiation and test temperature above are tabulated in Table II.
These critical fluences are valid only for the unfiltered 3200 °K color tempera-
ture spectrum utilized in these tests. It can be seen that while critical fluences
at 28 °C vary among cell types by an order of magnitude or more, at the lower
temperatures the critical fluences agree to within a factor of approximately two
or at most three.
3.3.2 Performance at Temperatures Other than that of Irradiation
The results described above apply to irradiations and characteriza-
tion measurements made without change in cell temperature. It was also
considered to be of some interest to observe the effect of irradiation at one
temperature upon performance at another temperature. For this purpose data
was taken for each cell at more than one temperature before each irradiation
series and again after the final irradiation step.
Tabulated in Table III are fractions of initial P remaining at
different temperatures after irradiation to 2.5 x 10 protons cm at 280C,
-70 "C and -160 °C. These data are from the cells also represented in
Figures 6 through 13. Similarities to the in-situ results can be noted: regard-
less of irradiation temperature, fractional performance loss is greatest at the
lowest operating temperature; differences between cell types is greatest when
the irradiation or the measurement temperature was 28 OC.
From the limited data obtained, it is not possible to determine a
correlation between the effects of identical irradiations at different tempera-
tures upon performance at some fixed temperature. Clearly degree of
16
Q N
0 00
0
a) x<
Co c
o ( (5
c - I I 3a D : wr
# fia '
· x o o <
CD
eltUt (s I)/ 3SI
1-6322
17
C-0i--
C:)U
c)CO
a,
CU
C'0
.
$4
'-4
o
orI-4
E
S0
0
01-4
U0
0
cO
C,)e
o
CQ
cDr
.r-
o-4
o
ell
o
CD
0
I:-
X I
m~~~~~~~~
CU~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C
li~~~~~~~~~~~~c.0~d X 0 I
CU~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C,
a)
'EI~~ ~ ~~~~~~~~~~~~~~~ I -4
C o
F:~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~b
I
/ x 0 0
I,-I.)
CD C,3/ U2CU
I~D~u(~si/ 0s
1-63fQ
o~~~~~~~~
Im~~lul Dss Dsb
I/o c6323: c
-4 -4 5
I I I0 CO CO 0~~~~~~~~P
-4~~a
WB!UIR4 OS ( ~)/OS
1- 6323o
18
N
0 c, Z o o ~1'~ 0
pq1.
z- Z -p4
Z0)
0) 0 c* C') CCo o o
_4 - C
1-4
X
C0
O
UOD 0
C.)Cl
I
cl
No
o O¢
S. bfl,q a)M Q
uCU)-4
o
0 -4 aco
bB'-
W)
o
Ielflul(OsI)/Os
1-6324
19
U
0cto
c-
.
4-
0)0)
EU
0
-4
0
P.'z
Q~~U ON
C) C3 . I-
0 0 0 0.4-4E E o C.t
CQ a I
c c
,o c -I
I .
C) C) -4,
1/1/ 11
/
I0-
CD
C)-4
0oo o
.-4
E 4-)
C CU
En
o a
o o-4
FX4r(W
O4o
TUIII(3s I)/ DSI
1-632520
P z zPc
aoa
; e
C) ')
P I C.)E
Eocu o
- 0V -4 =
C) a) 'O
CU CU M
* Xo
aa
" P4
Uo
cO
(4
C0C)
C)
")
:.v
rC
0
4Ii
o
04
1 2
-4 Uo
cq0
OCUE 4
o o n)
Pc
0-4
0)
.I~..4
01 -4
ie.ll.u(xeuc d )/ x e t
d
1-6326
21
C N
-4'1' <
cu
u
a,
C.1
00Cq
wCdli
:5a)CUC's
cUed
a)
r40:0
o
o C
PAp 4
xx z
aa)
E CuEoE
U I CUIC'
*0 -4rC: a
a1) a) 'CU·XO
* X 0
pZ0 0
5 5o UI IC:C
0
a) a)
CU C
0
-4
CD 0
· aa o
-4 -
u
U X
o x
0o
I'-
G!
(CD
CD
cl
04-
C's
Z S S
I I
ao C C
CC
.,4 .,CZ Q Q
0 0 4
I
x
x/ /
O4CO
C4-4
0 C
C)
0-.
o 4-
.0
p.4
o 0
ow-4
.,4
Pke
o
o
IU! u!(x udC )/x euci
1-6327
22
0
-4
o Q0o
.I
-4 44
o bD
O0 --
0 W$4
o~~~a
'M
a);.4
bfl[.L4
o0- 4
o
I *( Id )/ d
1-6328
23
u0
Co
I-4
C,ad
ad
C)
z
ad
-4
r.
cd
-4 -4
r0
Xad
P, pq
z Z
o i
E- C C o-4 -4
a a
* X" P-4
o x
Pq
z P 4 ZX - -
aZP
o C.) C.a a a
Cd) Cd) U)
uool~j
CO Z pq
S Ei
O O u
0 CO
Z
-4
QQC3
000
0o U .<00
CD
-4
0
CII
CC
GO-4 0o
CO
0 04-4
E m
o CU
o
b4
o1-4 a)
* ( d{)/' d1-6329
24
Oo0
I
-4
r--C
Cda1)
2U-
Cdr.0.- 41
CU
Cd
'.4r
pq.; -Ip p
Z ZX XC) Q .
- '0 0~0 -
04
* 3C
,5 X5 -4:D (D) X5
cl cl :
-E E .-~* X o:0 x 0
o
I
Table II. Critical 1 MeV Prot)gn Fluences. 2All entries are x 101 v protons/cm .
A. Fluence for 25% Decrease in Isc
Irradiation and Measurement Temperature
Cell Type i 28 °C -70 °C -160 °C -180 °C
Implanted 1l-cm Lopex N/P 2 x 1010 - 0.3 0.1Implanted 10-cm Lopex N/P 5 1 0.4Diffused 10-cm Lopex N/P 10 0.5 0.3 0.2Diffused 100'-cm CG N/P 8 - 0.3 0.2Diffused 10-cm CG P/N 1 0.3 0.1
B. Fluence for 50% Decrease in Isc
Irradiation and Measurement TemperatureCell Type 28 °C -70 °C -160 °C -180 °C
Implanted lQ-cm Lopex N/P 11 x 1010 2 0.9 0.3Implanted 10 n-cm Lopex N/P 23 4 2 -Diffused 100-cm Lopex N/P 40 3 0.7 0.7Diffused 10S-cm CG N/P 28 3 0.8 0.7Diffused 102-cm CG P/N 4 1 1 0.4
C. Fluence for 25% Decrease in Pmax
Irradiation and Measurement TemperatureCell Type 28 °C -70 °C -160 °C -180 °C
Implanted ln-cm Lopex N/P 1 x 1010 - 0.2 <0.1Implanted 10-cm Lopex N/P 3 0.7 0.4 -Diffused 10-cm Lopex N/P 6 0.4 0.2 0.2Diffused 10S-cm CG N/P 5 - 0.3 0.2Diffused 10-cm CG P/N 0.9 - 0.3 -
25
Table II. Continued.
D. Fluence for 50% Decrease in Pmax
Irradiation and Measurement TemperatureCell Type 28 °C -70 °C -160 °C -180 °C
Implanted IQ-cm Lopex N/P 6 x 101 0 2 0.6 0.3Implanted 10n-cm Lopex N/P 14 3 2 -Diffused 10f-cm Lopex N/P 24 2 0.6 0.5Diffused 10a-cm CG N/P 21 2 0.6 0.4Diffused 10Q-cm CG P/N 2 ,1 0.7 0.2
26
Table III. (Pmax)/(Pmax)initial versus Temperature AfterFixed Temperature Irradiations.
IrradiationCondition Cell Type 28°C -700C -160°C -180°C
2.5 x 1011 p/cm2 Implanted lQ•-cm Lopex N/P .28 .22 .15at 28 "C Implanted 10Q-cm Lopex N/P .38 .34 .27
Diffused 10'2-cm Lopex N/P .50 .41 .30Diffused 10f-cm CG N/P .45 .38 .37Diffused 10-cmn CG P/N .20 .22 .24
2.5 x 1011 p/cm2 Implanted ln-cm Lopex N/P .36 .23 .20at -70 °C Implanted 10-cm Lopex N/P .41 .29 .23
Diffused 10-cm Lopex N/P .33 .25 .21Diffused 102-cm CG N/P .38 .21 .18Diffused 10Q-cm CG P/N
2.5 x 1011 p/cm2 Implanted l1-cm Lopex N/P .39 .30 .17at -160 °C Implanted 10-cm Lopex N/P .51 .36 .21
Diffused 102-cm Lopex N/P .33 .25 .18Diffused 102-cm CG N/P .29 .27 .19Diffusedl10-cm CG P/N .32 .31 .26
27
(Pmax)/(Pmax)initial
degradation at a given measurement temperature is influenced by the irradiation
temperature. However a trend common to all the cell types is not observed.
For example, implanted 10 ohm-cm Lopex N/P cells showed degradation to 38%
of initial 28 °C 5.8 mW/cm2 maximum power when irradiated at 28 °C but only
to 51% of initial 28 °C 5.8 mW/cm2 maximum power when irradiated at -1 6 0 C.
Instead of showing greater fractional decrease under 28 °C irradiation, diffused
10 ohm-cm Lopex N/P cells degraded only to 50% of initial 28 °C 5.8 mW/cm2
output under 28 °C irradiation but to 33% of initial when irradiated at -160 °C.
It appears to be possible that after relatively high fluence irradiation levels,
P/N cells and implanted N/P cells may show less and diffused N/P cells more
output degradation at specific operating temperature as irradiation temperature
decreases. Further investigation may be warranted in this area.
3.3.3. Annealing Effects
The fact that cell performance at a given temperature after irradia-
tion is dependent upon temperature during irradiation indicates a temperature
dependence to the formation or stability of the radiation-induced crystal defects.
A possible distinct difference between the influence of the defects formed in
diffused and ion implanted cells suggests complexity beyond that which could be
investigated in this study. However a test was undertaken to determine whether
other parameters associated with a mission environment might also influence
degree of degradation of an irradiated cell.
A group of four N/P cells selected at random from available test11 2
lots were irradiated at -186 °C to 1.7 x 10 protons/cm . After measurement
the cells were held at -186 °C in the dark for 6 hours, under 5 mW/cm2 3200 °K
color temperature illumination for 5 hours then under 100 mW/cm2
illumination
for 6 hours. Finally the cells were warmed to 28 °C for 1 hour before returning
them to -186 °C for final measurements. The separate ultrastable tungsten
source used for cell measurements was carefully monitored to insure better
than 1% reproducibility over this period. An example of the sequence of cell
output measurements obtained is shown in Figure 14. It can be seen that modest
but significant performance recovery of this implanted 10 ohm-cm Lopex cell
28
O
co
-4
0o a)
LO 4j
C: 4-4~~~~~~~~~~
Ci2
c
o
C.)
-I
C
.-4
oo
I1 I-4 I%-J
,0 -o o CI~ 4.w-4-- ~~~~~~~~~~~
co ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~ - C d
.,,-
L,
cu~~~~~~c
co i1*1 CoC
q~ ~ ~~
-- ·~~~~~~~-E; C
cu a,~~~~~~~~~~~~~~~~~~~~~~~~~~a
· cu 031
C']~~~~~~~~o C 0 C']
(Vuw) I
1-6330
29
occurred during 6 hours of dark storage at -186 °C and little additional recovery
occurred during the next 5 hours under 5 mW/cm2 illumination. Six hours under
100 mW/cm2
illumination resulted in another step increase of output performance
and then warm-up to room temperature raised -186 C Pmax output to nearly
double the immediate post-irradiation level.
Normalized post-irradiation Isc and Pmax values for all four cells
after each storage cycle are tabulated in Table IV. It is to be noted that while
both of the implanted 10ohm-cmLopex cells showed some recovery under dark
storage, neither of the other cells exhibited any improvement until after the
5 hour 5 mW/cm storage. It is possible that the 'dark' anneal of the implanted
10 ohm-cm Lopex cells could actually have occurred during the 5 minute period
the cells were illuminated for measurement purposes. A definite improvement
occurred in all cells during storage under 100 mW/cm2 illumination.
From Table IV it is obvious that greatest fractional recovery of cell
performance produced by a temporary warm-up to room temperature. After
completion of irradiation and measurement, all of the cells in the in-situ tests
used to generate Figures 6 through 13 were allowed to warm up to room tem-
perature overnight and were remeasured at low temperatures a final time the
following day. In every case a performance improvement was observed. The
improvement was generally largest for the lowest temperature irradiations and
least but still significant in cells irradiated at 28 °C. Observed changes in
Pmax are given in Table V.
3.4 Effect of Irradiation on the Equivalent Circuit Model
Consideration of pre- and post-irradiation dark forward character-
istics allows the influence of radiation damage on the individual components of
the cell model to be determined. Figure 15 shows an example of 28 °C and
-183 °C dark forward characteristics of a typical cell before and after irradiation
at 28 °C. The 28 °C dark characteristic is substantially altered after irradiation
but the -183 °C characteristic is modified only slightly and only at current levels
exceeding a few milliamperes. This means that at room temperature there are
30
x.f4
.,q
11
L Cr O~ o C oO O~ O~C;C C'O , O o C- cO CD CO CO 00 LO
cqcccc q~~:cc cqcccc :;~~:c
00000 000 0 0001 0 0
N
.m'
N~E E~~ ,E hlEO
OO *O LO OO L.' LO L' L-- 0 ~ l L LOvv C.O O~ ' '.mlO "
E
Co. CO 0 , O,00 ! -"t co 0l
z~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
*s cs uW M LO rI 4uj L rI +jo '4- LQ urC OC n '- n r
.~ Eo o CDo o
c)~~~~~COL 31 oC 1 3 o 1 3 o
o I ot. . .I . Y . .. I .
o od C: LC
14 1. 5Hq sz5 q s s Hs 4$Hzs z z ss scq a c¢ cI ~ rD
0 $ k $ $ t $ $ t $ : 0 k ~ $ $ c 00r- Vq - -IT-
rc CS Cr) t U CV cO dt u: rc cu cv U: in H- CS cr, v U
o; o; E o;a
·C=~ C:
E- cZ r cscd cd c ; "a - l , P
0to
O0CoU 9
CDIo
'e E
o
U) C)
L)
O O
0,
eo
4e-
E-1
t~
O '
~u,-~
31
o
q>
0
C Cd
,-4,to a)
X O
p.I
P4R:
cu Vo o 4 d1 CO O
0 0 0 0C0oo c c oo 0 t 0
eCq CS d. .q:z 00 00 0 0 0
VU
0 C0 C C C0 C0
h,
k a, 0 0 0 0
U
Q ( O CA 4 0
cdd 0 0 0 0
I I aua I
0 O C14 Cd4 O- O O45j C) m c*: OD
0 0 0 ;U)
X UUC't o o
Cd CD C .CD CS C0 COQ )
U~
.
,.-.1 ,.-1 ( 0,
a-) 4 z P .4z 0
C C: co
ca) r ri rO
- - ifE~~CIQ)~~~~'*
32
U;
a)
0
p.4
Cd
Cd
0
~J
--
CTS
a)U
-1
- 2
103
10 -5
0.1 0.2 0.3 0.4 0.5 0.6
Vf (volt)
Figure 15.
0.7 0.8 0.9 1.0
The Effect of Irradiation on theDark I-V Characteristic.
1-6331
33
E
<l
basically two components to the degradation: (1) minority carrier collection effi-
ciency decreases because of increased recombination in the base, and (2) junction
saturation current increases because of the formation of generation-recombination
centers in the junction depletion region. Loss of collection efficiency appears as
decrease in output current while increased junction saturation current results in
a loss of output voltage. At low temperature and for low illumination intensities
cell degradation is essentially entirely due to loss of collection efficiency.
The example cell of Figure 15 was experimentally observed to de-
grade at 28 °C from initial ISC of 4.8 mA to 2.8 mA after irradiation and from
3.5 to 1.4 mA at -183 °C. From the forward curves it can be seen that a 90 mV
loss of V is expected at 28 °C while the -183 °C should be only 20 mV. Theseoc
values were experimentally verified.
Cell edge currents are of major consequence at low temperatures.
No changes in edge components have been observed. Consider Figure 16 which
shows an example cell with a severe edge channel double slope problem at
-184 C. Pre- and post-irradiation illuminated characteristics for this cell
were shown in Figure 4. The edge channel which has resistance of only 48 ohms
and dominates the characteristic between its turn-on at 0.4 volt and turn-on of
the cell junction just below 1 volt is observed to be totally unaffected by irradia-
tion to 1.7 x 1011 p/cm2 fluence.
The body component RJD of junction shunt resistance in a good cell
has no effect on the output I-V characteristics over the 28 °C and lower tempera-
ture range. It is possible that minor changes may occur in this component due
to irradiation but they are of no operational consequence. Data taken during
in-situ radiation tests did not resolve changes below 10-5 ampere associated
with this component in a good cell.
The equivalent circuit model of the cell is repeated in Figure 17 to
indicate radiation sensitivity of the circuit elements. A Schottky barrier in
series with the p-n junction has been omitted because although such a barrier
has previously been observed at low temperatures, its absence is now assured
in a cell properly fabricated for low temperature operation. The effect of irradia-
tion on this equivalent circuit model is basically the same as has been described
by Stofel and Joslin.(8)34
- I i I I I I i I I
Cell 10 CG 84-9Diffused h '' 10 O-cm CGIrradiatedc and Measuredat -184 C
After 1.7 x
\II
-47.6n RCH
= 35 KQ minimum
I I I I II I I I I I0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Vf (volt)
0.9 1.0
Figure 16. Dark I-V Characteristic of Excessive Edge Conduction Cell.
1-6332
35
10-1
10 - 2
1011
11
E-.
1-4
10- 3
10-4
10- 5
10- 6
0P-e
u1o
1-6333
36
C)
U)
C)
u
0m
U)la a)
c, c
O
o
Ca
0o
-4
,-4
CQCa
o
o
E05-4
-4
a)
0
0c4
C'
0u-
s-I
0C)
C)
. -
C)
0Z
C)
0Z
to
·.>, 0
-4 ct5UM °sz:(1) C
-CD
SECTION 4
CELL SPECIFICATION
Specification of design parameters for a solar cell to be employed
on a low temperature, low illumination intensity environment mission must be
based upon minimization of potential operational deficiencies due to fabrication
with simultaneous maximization of cell output and resistance to radiation-
induced degradation. While such a specification can easily be made, many
parameters must be selected almost arbitrarily and ultimate performance of
even the optimum cell will be drastically reduced by exposure to a high energy
proton radiation environment. Assuming that any solar cell being considered
will not exhibit serious initial performance deficiencies under low temperature
and intensity, true optimization can be approached by first considering the mode
of utilization required.
It is clear that regardless of the detail of its design parameters, any
silicon solar employed on a low temperature mission will be extremely vulner-
able to radiation degradation. All cell types tested have shown similar very
rapid degradation due to proton exposure at low temperatures. It has been
predicted( 4 ) that the radiation environment of Jupiter may include high fluxes
of protons with sufficient energies to penetrate even coverglasses as thick as
100 mils. In this case serious performance degradation will be incurred.
Large area arrays with 100 mil or thicker covers are probably not
a feasible approach to a solar cell power supply for a Jupiter flight. After
compensation for possible radiation induced degradation, a cell area exceeding2
4000 cm per watt would be necessary for -140 °C, 1/26 solar constant illumina-
tion operation in the vicinity of Jupiter.
A more practical solution could involve the use of small heavily
shielded cell arrays combined with solar concentrators to increase illumination
intensity on the panel area. Solar concentrators are not used for Earth orbit
applications because cell surface intensities greater than one solar constant
37
result in heating and series resistance problems with overall decrease in per-
formance-weight-cost optimization. However at low temperature under 1/26
solar constant illumination, a concentration factor exceeding 10 could be em-
ployed without heating or series resistance problem. In this case a cell area
of 150 cm2 or less per watt could be sufficient. The small area array would be
shielded with as much as 100 mils of silica coverplate to protect against protons
with energies up to 22 MeV. As highest cell efficiency is achieved at lowest
temperatures the cell array and its supports would be designed for maximum
heat removal. A small panel/concentrator approach in addition to offering the
possibility for adequately protecting the cells without excessive weight penalties
would also provide a coincidental improvement in cell output. Minor shunt or
double slope problems which might be significant at 1/26 solar constant become
negligible at decade higher current levels.
4.1 Cell Specification
A cell optimized for use in a low temperature/low intensity environ-
ment can be similar to existing conventional cell structures. Provision must
be made to assure absence of low temperature performance problems. The
parameters below are recommended for cells to be utilized in a small panel/
concentrator configuration with thick cover protection over cells. IPC considers
these to define a best choice for the optimized cell. However, indicated param-
eters are not necessarily critical and as will be briefly discussed below,
acceptable alternatives could be employed.
1. Cell Structure: N+/PP+
2. Base Material: Low oxygen/low dislocation
3. Base Resistivity: 10 - 30 ohm-cm
4. Cell Thickness: 8 - 10 mils
5. Front Surface Finish: Polish etched to mirror surface
6. Junction Formation: Diffusion or implantation
38
7. Junction Depth:
8. Back surface P+Layer:
9. Contact Material:
10. Front Conl;tactConfiguration*:
11. Series resistance*:
12. Shunt Resistance:
1 3. AntireflectiveCoating:
14. Cell SurfaceDimensions:
15. Method of CellSizing:
16. Cover Protection*:
17. Interconnections:
Alloyed aluminum
Aluminum or titanium- silver
Standard 5 through 14 fingers pluscontact bar. Up to 6% of activesurface occupied.
< 0.3 ohms (2 x 2 cm cell)
2000 ohms minimum
CeO2 or TiOx
applied before cell
is sized
2 cmx2 cm
Scribing after contacting and ARcoating
Glued fused silica as thick as possi-ble up to 100 mils
Ultrasonically welded aluminum orsoldered
Specification of an N+/PP+ cell structure is straightforward. P/N
cells are considerably more sensitive to radiation damage at 28 °C than are
N/P's and are definitely not superior to N/P's under low temperature irradiations.
P/N's also show inferior performance at low temperature after being irradiated
at room temperature and the choice of N/P follows. A back surface P+ layer
*These recommendations apply specifically to the case of a heavily protectedpanel used in conjunction with a solar concentrator to increase intensity by afactor of approximately ten.
39
0.3 - 0.5 micron
necessary to eliminate Schottky-barrier rectification effects at low temperature
completes the N+/PP+ structure.
Investigations conducted elsewhere( 2 ) have shown that cells made of
crucible grown silicon exhibit greater loss of current with decreasing tempera-
ture than do cells made of low oxygen content material. As the crucible grown
material cells do not exhibit any compensating improved performance or radia-
tion resistance characteristics at low temperatures, choice of the low oxygen,
low dislocation base material is also clear. Cells with 10-cm base resistivity
showed slight overall radiation resistance superiority over 100-cm base cells.
As at reduced temperatures there exists little dependence of cell Voc on base
resistivity the 10-cm cells offers no advantageous initial performance. Optimum
base resistivity is then at least 10 ohm-cm and could possibly be as high as 100
ohm-cm. In the absence of test data on 100 ohm-cm cells, specification not
greatly above 10 ohm-cm is prudent.
Designation of cell thickness as 8 - 10 mils is somewhat arbitrary.
Increasing cell thickness to more than twice the base minority carrier diffusion
length results in negligible increase of output. At -140 °C a cell of 8 - 10 mils
thickness should give maximum response.
The low temperature/intensity cell must have good junction char-
acteristics. In order to insure low leakage it is possible to use a deep junction
or a normally shallow junction with good uniformity characteristics. As a deep
junction (- 1 micron) results in noticeable loss of blue response, it is preferable
to use a 0.3 to 0.5 micron junction introduced into a mirror finish surface with-
out visible scratches or nonuniformities. The 0.3 to 0.5 micron depth is also
somewhat deeper than normal but is required to give good results with aluminum
contacts. A shallower 0.25 micron junction could be acceptable for titanium-
silver contacts. Equally acceptable performance is observed with implanted
and diffused junction cells.
The back surface P+ layer below the contact metallization can be
produced either by alloying aluminum into the silicon or by a deep diffusion.
Alloyed aluminum is considered to be superior because it results in a layer
40
which is reflective to unabsorbed photons and consequently yields slight improve-
ment in long wavelength response.
Optimized front contact configuration is dependent upon the method
of cell utilization. A 2 x 2 cm cell operating under 1/26 solar constant illumina-
tion requires only a contact bar and perhaps two fingers to keep its series
resistance to an acceptable level which does not have to be less than one ohm.
However if a solar concentrator is used to increase cell surface illumination
intensity to a level approaching one solar constant, contact configurations
similar to those currently in use will be required. A 2 x 2 cm cell operating
under 50 mW/cm2
or greater illumination should have series resistance not
greater than 0.3 ohm.
Antireflective coatings presently in use will be adequate for the low
temperature cell. Choice of CeO2 or TiOx is made because of their better
optical optimization for a covered cell but SiOx
is also acceptable. IPC con-
siders it to be possible that the presence of antireflective coating over the
cell edge could contribute to edge inversion effects and could increase the double
slope edge conduction problem. No attempt has been made to investigate this
possibility. If the cell is sized by scribing after contacts and AR coating are
applied the probability of encountering an edge conduction problem is low.
Thickness of the cover protection is particularly important. In
the absence of experimental data or reliable theory to characterize the trapped
proton environment of Jupiter, it is recommended that 100 mils of fused silica
be used to provide protection against energies up to 22 MeV. This recommenda-
tion should be modified by latest available information should design of a solar
power supply actually be undertaken. It is also of considerable importance that
equivalent thick protection be provided to the contact bar and cell back areas as
well as to the active surface.
41
SECTION 5
TEST PLAN
Following fabrication of quantities of cells to be utilized on a low
temperature/low intensity mission it would be necessary to eliminate any cells
showing significant performance deficiencies. Observation of the dark forward
I-V characteristic at -196 °C during direct submersion in liquid nitrogen will
identify any cell with a Schottky-barrier contact problem, a low shunt resistance
or a double slope problem. Only a low short-circuit current because of non-
linear dependence of Isc on temperature is not observed by this test. This
non-linear current problem has not been associated with low oxygen content
(Lopex) silicon cells but even for these types it would be desirable to make an
illuminated measurement at -135 °C or -196 °C on a sample cell from each
crystal used to confirm acceptable Isc.
The simplified low temperature dark forward measurement is made
using mechanical-electrical clip connections to the cell which is submerged into
a dark liquid nitrogen bath. The forward I-V characteristic is displayed on a
curve tracer oscilloscope. A linear display covering 0-10 mA and 0-1.2 volt is
adequate. A good 2 x 2 cm cell will exhibit a characteristic similar to that in
the upper display of Figure 18 meeting the following conditions:
(1) forward current < 0.5 mA at 0.90 volt
(2) forward current > 10 mA at 1.1 volt
A cell with low shunt resistance or double slope problem will violate require-
ment (1) while a cell with rectifying contact will not meet requirement (2).
A cell which passes the above acceptance test could still have low
short circuit current at -135 °C. If the cell is fabricated from low oxygen float
zone or Lopex type material, low Isc is not expected. But because the mechanism
for non-linear dependence of Isc on temperature has not been determined it
could be advisable to make a low temperature/low intensity output measurement
on at least one sample cell from each silicon crystal lot. A single measurement
42
0,1 V/div _-
Ia. Typical Good Cell
b. Cell with Non-Ohmic RearContact
c. Cell with Ex-cessive Con-duction at0.8 volt
Figure 18. Representative Dark Linear I-V Characteristics at 77 OK.
1-6001
43
1 mA/div
at -135 °C or -196 °C under low intensity AMO spectrum illumination should
confirm acceptability of cells from the same crystal. An acceptable cell will
test as follows:
Minimum Isc 0.22 (mA/cm 2)/(mW/cm ) at -135 °C
Minimum Isc 0.20 (mA/cm2)/(mW/cm2 ) at -196°C
44
SECTION 6
REFERENC ES
1. Kirkpatrick, A. Ro, Ho, J. C. and Bartels, F. T. C., 'Analysis ReportVolume I', Contract NAS2-5516.
2. Brandhorst, H. W. and Hart, R. E., 'Spectral Responce of Silicon SolarCells at Low Temperatures', p. 142, Conference Record of Eighth IEEEPhotovoltaic Specialists Conference, Seattle, August 1970.
3. Ho, J. C., Bartels, F. T. C. and Kirkpatrick, A. R., 'Solar Cell LowTemperature, Low Solar Intensity Operation', p. 150, Conference Recordof Eighth IEEE Photovoltaic Specialists Conference, Seattle, August 1970.
4. Haffner, J. W., 'Calculated Dose Rates in Jupiter's Van Allen Belts',Paper 69-18, AIAA 7th Aerospace Sciences Meeting, New York, January1969.
5. Haffner, J. W., 'Natural Nuclear Radiation Environments for the GrandTour Missions', IEEE Trans. Nuc. Science, NS-18, No. 6, 443, December1971.
6. Wilsey, N. D. and Lambert, R. J., 'Low Temperature Irradiations ofSilicon Solar Cells with 1 MeV Electrons', p. 169, Conference Record ofEighth IEEE Photovoltaic Specialists Conference, Seattle, August 1970.
7. Debs, R. J. and Hanes, N. R., 'Preliminary Results of Radiation andJupiter Environment Tests on Solar Cells', p. 155, Conference Recordof Eighth IEEE Photovoltaic Specialists Conference, Seattle, August 1970.
8. Stofel, E. and Joslin, D., 'Low-Energy Proton Damage to Silicon SolarCells', IEEE Trans. Nuc. Science, NS-17, No. 6, 250, December 1970.
45